[COFF] Happy birthday, John Backus!
Clem Cole
clemc at ccc.com
Thu Dec 5 01:08:37 AEST 2019
BTW: we called that test snippet "Eklund.f" in honor of Dave Eklund who was
the 'main man' behind DEC and Intel's Fortran compilers for many, many
years (he finally retired a few years back). Among other things, Dave was
always looking for interesting fixed format test cases. To be fair, when
I first tripped over the code snippet around the time Dave was retiring, I
sent it to him (and the rest of the Fortran compiler team). The "HP"
Alpha/Itanium compilers (*ie.* the DEC GEM based ones) were fine with it.
At the time, the Intel compiler accepted the deck without error but tossed
cookies in the runtime. Lori Menard and I filled a bug report and she
added it to the Intel Fortran regression test suite.
Here is a version of the same deck that is a tad more intelligible and more
likely to be accepted by lessor compiler front-ends:
PROGRAM BLOCK
INTEGER I1,I2,I3,I4,I5
DIMENSION I1(13),I2(13)
I4=1
I5=2
I3=13
CALL PASCAL(I1,I2,I3,I4,I5)
END
SUBROUTINE PASCAL(IP1,IP2,IP3,IP4,IP5)
INTEGER IP3,IP1(IP3),IP2(IP3),IP4,IP5
INTEGER IP6,IP7
DO IP6=IP4,IP3
DO IP7=IP5,IP6-IP4
IP1(IP7)=IP2(IP7)+IP2(IP7-IP4)
ENDDO
IP1(IP4)=IP4
IP1(IP6)=IP4
DO IP7=IP4,IP3-IP6
WRITE(*,'(" "$)')
ENDDO
DO IP7=IP4,IP6
WRITE(*,'(I4$)') IP1(IP7)
ENDDO
WRITE(*,*)
DO IP7=IP4,IP3
IP2(IP7)=IP1(IP7)
ENDDO
ENDDO
END
On Wed, Dec 4, 2019 at 9:54 AM Clem Cole <clemc at ccc.com> wrote:
> Intel ifort <https://software.intel.com/en-us/fortran-compilers> which as
> one of my friends like to say "has the DEC Fortran compiler DNA ground up
> and reinserted" ;-)
>
> On Wed, Dec 4, 2019 at 12:41 AM Wesley Parish <wobblygong at gmail.com>
> wrote:
>
>> I've just run that little program through gfortran without switches
>> (naming it pyramid.for) and it comes up with:
>>
>> pyramid.for:25:6:
>>
>> ;ENDDO;IMPLICIT(IF)=IF;IMPLICIT(EXTERNAL)=IF;DOFORRETURN=IF,GOTO-EX
>> 1
>> Error: Bad continuation line at (1)
>> pyramid.for:12:9:
>>
>> R;DOFORIF=INTEGER,INTEGER;ENDDO;INTEGER=IF+IF;GOTO=INTEGER*INTEGER*
>> 1
>> Warning: Deleted feature: Loop variable at (1) must be integer
>> pyramid.for:23:56:
>>
>> IF,INTEGER,EXTERNAL,RETURN;DOFOREXTERNAL=IF,GOTO;DOFORRETURN=INTEGE
>> 1
>> Warning: Deleted feature: Loop variable at (1) must be integer
>> pyramid.for:25:52:
>>
>> ;ENDDO;IMPLICIT(IF)=IF;IMPLICIT(EXTERNAL)=IF;DOFORRETURN=IF,GOTO-EX
>> 1
>> Warning: Deleted feature: Loop variable at (1) must be integer
>> pyramid.for:26:43:
>>
>> XTERNAL;WRITE(IF,'(''$ '')');ENDDO;DOFORRETURN=IF,EXTERNAL;WRITE(I
>> 1
>> Warning: Deleted feature: Loop variable at (1) must be integer
>> pyramid.for:27:62:
>>
>> IF,'(''$''I4)')IMPLICIT(RETURN);ENDDO;WRITE(IF,'( /)');DOFORRETURN=
>> 1
>> Warning: Deleted feature: Loop variable at (1) must be integer
>> pyramid.for:23:34:
>>
>> IF,INTEGER,EXTERNAL,RETURN;DOFOREXTERNAL=IF,GOTO;DOFORRETURN=INTEGE
>> 1
>> Warning: Deleted feature: Loop variable at (1) must be integer
>> pyramid.for:15:17:
>>
>> LSUBROUTINE(IMPLICIT,LOGICAL,GOTO,IF,INTEGER)
>> 1
>> Warning: Rank mismatch in argument ‘implicit’ at (1) (rank-1 and
>> scalar) [-Wargument-mismatch]
>>
>> I think gfortran must have '--strict' on by default. I might try it
>> with the OpenWatcom Fortran compiler next - if it'll install on my
>> current Linux box.
>>
>> Just out of interest, what is your current Fortran 2018 compiler?
>>
>> Wesley Parish
>>
>> On 12/4/19, Clem Cole <clemc at ccc.com> wrote:
>> > On Tue, Dec 3, 2019 at 2:56 AM Dave Horsfall <dave at horsfall.org> wrote:
>> >
>> >> As every computer programmer should know, John Backus was emitted in
>> >> 1924;
>> >> he
>> >> gave us the BNF syntax, but the sod also gave us that FORTRAN
>> >> obscenity...
>> >>
>> > Be careful, Fortran still pays a lot of bills (I like to say that it has
>> > paid my salary for nearly 45 years and I don't program in it - I'm an OS
>> > guy). But Fortran >>is<< the #1 language for anything scientific and I
>> > don't think that is going away in the future or really change its
>> position
>> > in popularity for a number of reasons (my analog is the QWERTY keyboard
>> -
>> > that ship has sailed and it's not economically interesting). There are a
>> > number of places to check this out, but try looking at Archer AC Code
>> > Status
>> > <http://www.archer.ac.uk/status/codes/>, which is an interesting HPC
>> usage
>> > site in the UK. Note that Fortran is by far the leading programming
>> > language used for ‘production’ (there are other sites that offer similar
>> > data, I'll leave it to the reader to find them).
>> >
>> > Trivia: there is no way that FORTRAN can be described in any syntax; it
>> is
>> >> completely ad-hoc.
>> >
>> > Again, be careful with such observations. First off, I'm fairly sure
>> that
>> > the Intel Compiler teams (ifort
>> > <https://software.intel.com/en-us/fortran-compilers>) use a parser
>> > generator for parts of the ifort front-end. (Paul W might know more
>> details
>> > as he once worked in that technology). As I understand it, the
>> front-end
>> > does have a number of special cases in it, so your observation is
>> partially
>> > true, but the language definition is not 'completely ad-hoc'.
>> >
>> > The facts are that the language my father learned in the early 1960s
>> > (FORTRAN-II) and the language I learned in the late 60's/early 1970s
>> > (FORTRAN-IV) are not the same language as today's Fortran-2018, i.e. the
>> > language definition has hardly been static. Said in another way, about a
>> > year ago, a new standard for Fortran 2018 standard was released – see
>> > Fortran
>> > 2018 (Formerly Fortran 2015) <https://wg5-fortran.org/f2018.html> and
>> > it actually
>> > offers support for ‘modern’ ideas such as object-oriented programming:
>> > Object-oriented
>> > programming in Fortran Wiki
>> > <http://fortranwiki.org/fortran/show/Object-oriented+programming>
>> > .
>> >
>> > FWIW: I tried to explain some of these ideas pictorially in my Quora
>> > answer: Clem Cole's answer to Is Fortran obsolete?
>> > <https://www.quora.com/Is-Fortran-obsolete/answer/Clem-Cole> This is
>> not to
>> > denigrate other languages like Julia, Tensorflow etc. But the fact is
>> that
>> > the hammer has been improved and the *ways the nails are delivered has
>> > changed*, but the *fundamental action provided* (fastening for nails and
>> > scientific computation in the case of Fortran) has been unchanged
>> because
>> > it has proven to be the one of the best, if not the best to do the job
>> it
>> > is designed to do.
>> >
>> > That said, I offer the following code snippet, which my Intel 2018
>> > compatible compiler accepts without any switches. Which is really a
>> > remarkable comment about the women and men in the front-end team:
>> >
>> > C This FORTRAN program may be compiled and run on a Norsk Data
>> > C computer running SINTRAN and the FTN compiler. It uses only
>> > C FORTRAN reserved words, and contains just one numerical
>> > C constant, in a character string (a format specifier). When
>> > C you run it, it prints a well known mathematical construct...
>> > C
>> > C Even FORTRAN is a block structured programming language:
>> > C
>> > PROGRAM
>> > ;PROGRAM;INTEGERIF,INTEGER,GOTO,IMPLICIT;REALREAL,DIMENSION,EXTERNA
>> > AL,FORMAT,END;INTEGERLOGICAL;REALCOMPLEX,DATA,CALL,ASSIGN,CHARACTER
>> > R;DOFORIF=INTEGER,INTEGER;ENDDO;INTEGER=IF+IF;GOTO=INTEGER*INTEGER*
>> > *INTEGER*INTEGER-INTEGER-IF;CALLFUNCTION(IMPLICIT,REAL,DIMENSION,EX
>> > XTERNAL,FORMAT,END,LOGICAL,COMPLEX,DATA,CALL,ASSIGN,CHARACTER);CALL
>> > LSUBROUTINE(IMPLICIT,LOGICAL,GOTO,IF,INTEGER)
>> > END
>> > SUBROUTINEFUNCTIO
>> > ON(IMPLICIT,REAL,DIMENSION,EXTERNAL,FORMAT,END,LOGICAL,COMPLEX,DATA
>> > A,CALL,ASSIGN,CHARACTER);RETURN
>> > END
>> > SUBROUTINESUBROUTINE(IMPLICIT,L
>> > LOGICAL,GOTO,IF,INTEGER);INTEGERGOTO,IMPLICIT(GOTO),LOGICAL(GOTO),I
>> > IF,INTEGER,EXTERNAL,RETURN;DOFOREXTERNAL=IF,GOTO;DOFORRETURN=INTEGE
>> > ER,EXTERNAL-IF;IMPLICIT(RETURN)=LOGICAL(RETURN)+LOGICAL(RETURN-IF);
>> > ;ENDDO;IMPLICIT(IF)=IF;IMPLICIT(EXTERNAL)=IF;DOFORRETURN=IF,GOTO-EX
>> > XTERNAL;WRITE(IF,'(''$ '')');ENDDO;DOFORRETURN=IF,EXTERNAL;WRITE(I
>> > IF,'(''$''I4)')IMPLICIT(RETURN);ENDDO;WRITE(IF,'( /)');DOFORRETURN=
>> > =IF,GOTO;LOGICAL(RETURN)=IMPLICIT(RETURN);ENDDO;ENDDO
>> > END
>> > Running the program should yield:
>> > 1
>> > 1 1
>> > 1 2 1
>> > 1 3 3 1
>> > 1 4 6 4 1
>> > 1 5 10 10 5 1
>> > 1 6 15 20 15 6 1
>> > 1 7 21 35 35 21 7 1
>> > 1 8 28 56 70 56 28 8 1
>> > 1 9 36 84 126 126 84 36 9 1
>> > 1 10 45 120 210 252 210 120 45 10 1
>> > 1 11 55 165 330 462 462 330 165 55 11 1
>> > 1 12 66 220 495 792 924 792 495 220 66 12 1
>> >
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/coff/attachments/20191204/daa8571c/attachment.html>
More information about the COFF
mailing list